China FAW Group Co., Ltd. (FAW for short) can be traced back to China First Automobile Works. After more than 60 years of development, FAW has grown into a large state-owned automobile enterprise group with an annual production and sales volume of 3 million vehicles. Its total production and sales volume has always ranked among the first echelon in the industry. FAW shoulders the responsibility of building China into an "automotive power". Considering that it is an inevitable route to drive industrial chain upgrading with digitalization, FAW pioneered digital transformation. In 2018, FAW took the lead in adopting the public cloud HPC cluster of eSurfing Cloud, to drive the transformation of the automotive industry from traditional manufacturing to smart manufacturing with cloud technologies.

For an automotive company, vehicle safety is the most critical aspect and a link to which every auto company attaches great importance. A mature vehicle needs 60 crash tests of the real vehicle before it goes on the market, supported by 300 simulation tests. Under the traditional computing mode, one simulated crash test takes at least 30 hours, and 300 simulated tests require 3-4 years to complete. Therefore, the launch process of a new model, i.e. design-testing-production-consumer, takes at least 5-6 years. eSurfing Cloud builds an exclusive HPC cluster for FAW that provides one-stop response to HPC resource requirements in crash simulations during automotive R&D. The solution reduces the time of a single simulation experiment from 8 hours to 2.5 hours and shortens the R&D cycle by 60%. The overall computing power increases by 30%, which greatly boosts the development of FAW's smart manufacturing.

Through digital R&D, FAW has created a more agile vehicle development model. In order to meet the top quality standards of new Hongqi series, FAW has established a world-leading R&D system, and continuously accelerates independent R&D and technological innovations and shortens the R&D cycle. Wherein, the local IDC of FAW has built its own HPC cluster with thousands of cores to meet the CAE simulation design needs in automotive R&D. Computing demand at peak hours is often 4-5 times higher than usual, so computing tasks are often postponed to a week later, which seriously affects R&D efficiency and hinders fast product iteration to a certain extent. If the traditional model is adopted and computing power is configured subject to peak hours, the resources will be idle at certain time periods. Meanwhile, the traditional model has a long construction period, which cannot quickly meet the needs of emergent services.

Leveraging on the cloud platform, dedicated line and other resource capabilities of eSurfing Cloud, FAW builds a hybrid cloud architecture, under which the two centers can both carry computing loads and share service traffic during peak business hours. The public cloud HPC resources with thousands of cores can shorten the total operation time by 30% to 40%, thereby improving R&D efficiency and accelerating vehicle model iteration. This can help the company to grab an upper hand in market competition.